Stable carbon and nitrogen isotopic composition of leaves, litter, andsoils of various ecosystems along an elevational and land-usegradient at Mount Kilimanjaro, TanzaniaFriederike Gerschlauer1, Gustavo Saiz2,1, David Schellenberger Costa3, Michael Kleyer3, Michael Dannenmann1, andRalf Kiese1
1Institute of Meteorology and Climate Research, Karlsruhe Institute of Technology, Garmisch-Partenkirchen, Germany2Department of Environmental Chemistry, Faculty of Sciences, Universidad Católica de la Santísima Concepción,Concepción, Chile3Department of Biology and Environmental Sciences, University of Oldenburg, Oldenburg, Germany
Received: 14 September 2018 – Discussion started: 22 October 2018Revised: 28 December 2018 – Accepted: 13 January 2019 – Published: 25 January 2019
Abstract. Variations in the stable isotopic composition ofcarbon (δ13C) and nitrogen (δ15N) of fresh leaves, litter,and topsoils were used to characterize soil organic mat-ter dynamics of 12 tropical ecosystems in the Mount Kil-imanjaro region, Tanzania. We studied a total of 60 sitesdistributed along five individual elevational transects (860–4550 m a.s.l.), which define a strong climatic and land-usegradient encompassing semi-natural and managed ecosys-tems. The combined effects of contrasting environmentalconditions, vegetation, soil, and management practices hada strong impact on the δ13C and δ15N values observed in thedifferent ecosystems. The relative abundance of C3 and C4plants greatly determined the δ13C of a given ecosystem. Incontrast, δ15N values were largely controlled by land-use in-tensification and climatic conditions.
The large δ13C enrichment factors (δ13Clitter− δ13Csoil)
and low soil C/N ratios observed in managed and disturbedsystems agree well with the notion of altered SOM dynam-ics. Besides the systematic removal of the plant biomasscharacteristic of agricultural systems, annual litterfall pat-terns may also explain the comparatively lower contents ofC and N observed in the topsoils of these intensively man-aged sites. Both δ15N values and calculated δ15N-based en-richment factors (δ15Nlitter− δ
15Nsoil) suggest the tightest ni-trogen cycling at high-elevation (>3000 m a.s.l.) ecosystemsand more open nitrogen cycling both in grass-dominatedand intensively managed cropping systems. However, claimsabout the nature of the N cycle (i.e. open or closed) should
not be made solely on the basis of soil δ15N as other pro-cesses that barely discriminate against 15N (i.e. soil nitrateleaching) have been shown to be quite significant in MountKilimanjaro’s forest ecosystems. The negative correlation ofδ15N values with soil nitrogen content and the positive corre-lation with mean annual temperature suggest reduced miner-alization rates and thus limited nitrogen availability, at leastin high-elevation ecosystems. By contrast, intensively man-aged systems are characterized by lower soil nitrogen con-tents and warmer conditions, leading together with nitrogenfertilizer inputs to lower nitrogen retention and thus signifi-cantly higher soil δ15N values. A simple function driven bysoil nitrogen content and mean annual temperature explained68 % of the variability in soil δ15N values across all sites.Based on our results, we suggest that in addition to land-useintensification, increasing temperatures in a changing climatemay promote soil carbon and nitrogen losses, thus alteringthe otherwise stable soil organic matter dynamics of MountKilimanjaro’s forest ecosystems.
Published by Copernicus Publications on behalf of the European Geosciences Union.
410 F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro
1 Introduction
Conversion of natural ecosystems to agriculture is a world-wide phenomenon, which is of particular significance in trop-ical regions where human population growth rates are cur-rently the highest (FAO and JRC, 2012). Changes in cli-mate and land use significantly alter vegetation compositionand biogeochemical cycles, causing a strong impact on car-bon (C) and nitrogen (N) turnover and stocks (Smith et al.,2014). Tropical forest biomes are particularly relevant in thiscontext, as they are significant C storages and N turnoverhotspots (Bai et al., 2012; Hedin et al., 2009; Lewis et al.,2009; Pan et al., 2011; Vitousek, 1984). Considering the in-creasing pressure on natural land, it is becoming even morecrucial to understand how anthropogenic interventions affectecosystem C and N cycling and gain better knowledge aboutthe main drivers of nutrient cycling, as well as the associatedexchange processes with the atmosphere and hydrosphere intropical environments.
Research exploiting the natural abundance of stable iso-topes has proved quite suitable for investigating potential im-pacts of land-use and/or climate change on C and N cyclingin terrestrial systems (Michener and Lajtha, 2007; Panettieriet al., 2017; Saiz et al., 2015a). Variations in the stable iso-topic composition of C (δ13C) and N (δ15N) in plants andsoils are the result of fractionation processes occurring dur-ing ecosystem exchange of C and N. Thus, δ13C and δ15Ncan serve as valuable indicators about ecosystem state andprovide useful insights on how these systems respond to bi-otic and abiotic factors (Dawson et al., 2002; Högberg, 1997;Ma et al., 2012; Pardo and Nadelhoffer, 2010; Peterson andFry, 1987; Robinson, 2001).
Plants discriminate against 13CO2 (carbon dioxide) duringphotosynthetic CO2 fixation depending on plant metabolism(i.e. C3 and C4 photosynthetic pathways). Most tropicalgrasses typically employ the C4 photosynthetic pathway(δ13C values > − 15 ‰), while trees and shrubs use the C3photosynthetic pathway (δ13C values< −24 ‰) (Bird et al.,1994; Bird and Pousai, 1997; Cernusak et al., 2013; Far-quhar et al., 1980). The distribution of C3 and C4 vegeta-tion shows clear patterns along elevational gradients, withincreasing abundance of C3 species towards high elevations(Bird et al., 1994; Körner et al., 1991; Tieszen et al., 1979).Environmental conditions such as water availability also ex-ert a significant influence on isotopic discrimination duringatmospheric CO2 fixation. Accordingly, compared to opti-mal moisture conditions, water stress leads to the enrichmentof 13C in C3 plants (Farquhar and Sharkey, 1982), while thisisotopic fractionation is less obvious or even absent in C4plants (Ma et al., 2012; Swap et al., 2004).
The soil organic matter (SOM) pool integrates the iso-topic signature of the precursor biomass over different spatio-temporal scales (Saiz et al., 2015a). Variation in soil δ13Cvalues represents a valuable tool to better assess SOM dy-namics, mineralization processes, or reconstruct past fire
regimes (Saiz et al., 2015a; Wynn and Bird, 2007). The δ13Cof SOM in a given ecosystem is greatly controlled by therelative abundance of C3 and C4 plants due to their contrast-ing C isotopic composition. Therefore, strong variations insoil δ13C can also be used to identify sources of particulateorganic matter and vegetation shifts such as woody thicken-ing. However, fractionation effects associated with differen-tial stabilization of SOM compounds, microbial reprocessingof SOM, soil physico-chemical characteristics, and the ter-restrial Seuss effect preclude a straightforward interpretationof soil δ13C values (Saiz et al., 2015a).
Plant and soil δ15N relate to environmental and man-agement conditions controlling N turnover, availability, andlosses. δ15N values in soils are generally more positive thanthose of vegetation due to the relatively large isotopic frac-tionation occurring during soil N transformations (Dawsonet al., 2002). The N cycle of a given ecosystem may be char-acterized as closed if both efficient microbial N retention andthe absence of external N inputs (e.g. atmospheric depositionand fertilizer additions) prevent substantial gaseous and/orleaching N losses. In contrast, open ecosystem N cycling ischaracterized by significant inputs and losses of N. On theone hand, gaseous N losses from soils are strongly depletedin 15N due to the high fractionation factors associated withthese processes (Denk et al., 2017). This results in high δ15Nvalues of the residual substrate, which consequently leavesless importance to impacts of external N additions (Robin-son, 2001; Zech et al., 2011). On the other hand, N leach-ing seems to only discriminate slightly against ecosystem15N. According to Houlton and Bai (2009) δ15N values ofdrained water agree well with those of soils across variousnatural ecosystems worldwide. Moreover, it is also importantto consider the possibility that soil δ15N may also be influ-enced by other factors including rooting depth, uptake of dif-ferent N compounds, and symbiotic N2 fixation (Nardoto etal., 2014). Variations in δ15N values of plants and soils havebeen successfully applied to characterize N cycling acrossa large variety of ecosystems worldwide (Amundson et al.,2003; Booth et al., 2005; Craine et al., 2015a, b; Martinelli etal., 1999; Nardoto et al., 2014). This includes research workthat has particularly focused on the study of N losses derivedfrom land-use changes or intensification (Eshetu and Hög-berg, 2000; Piccolo et al., 1996; Zech et al., 2011).
Information on ecosystem C and N cycling is still scarcein many tropical ecosystems, particularly in remote regionsof Africa (Abaker et al., 2016; 2018; Saiz et al., 2012;Townsend et al., 2011). Furthermore, feedbacks between Cand N cycles, such as limitations of N availability in ecosys-tem C sequestration and net primary productivity in trop-ical forest, require urgent investigations (Gruber and Gal-loway, 2008; Zaehle, 2013). In such a context, the Kiliman-jaro region in Tanzania offers the rare possibility to study abroad range of tropical ecosystems across contrasting land-use management intensities and varying climatic conditions.This region hosts a large variety of semi-natural and managed
F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro 411
ecosystems as a result of the strong elevational and land-usegradient.
We hypothesized that (i) vegetation composition (C3 /C4)
is the main control for ecosystem δ13C values, whereas(ii) δ15N values are rather controlled by land-use manage-ment and climatic conditions. The main aim of this study isto evaluate the potential of δ13C and δ15N values in plant andsoil material to assess C and N cycling across a broad vari-ety of semi-natural and managed ecosystems under varyingclimatic conditions.
2 Materials and methods
2.1 Study sites
This study was conducted on the southern slopes of MountKilimanjaro (3.07◦ S, 37.35◦ E; 5895 m a.s.l.) in north-eastTanzania. The climate is characterized by a bimodal precipi-tation pattern with a major rainy season between March andMay and the other peak between October and November. Re-cently, Appelhans et al. (2016) used a network of 52 meteo-rological stations strategically deployed in the Kilimanjaroregion to measure air temperature and precipitation. Theythen used geo-statistical and machine-learning techniques forthe gap filling of the recorded meteorological time series andtheir regionalization, which provides the means to calculatethe meteorological data used for the complete set of sites (60)used in our work. Please refer to Appelhans et al. (2016) formore details. Maximum mean annual precipitation (MAP)of 2552 mm occurs at an elevation of around 2260 m a.s.l.,decreasing towards lower and higher elevations, reaching657 and 1208 mm yr−1 at 871 and 4550 m, respectively (Ta-ble 1). Variations in air temperature are dominated by diurnalrather than seasonal patterns (Duane et al., 2008). Mean an-nual temperature (MAT) decreases with increasing elevation,ranging from 24.8 ◦C at 860 m to 3.5 ◦C at 4550 m (Table 1).
Five altitudinal transects ranging from 860 to 4550 m a.s.l.were established along the mountain slopes. At each tran-sect, 12 ecosystems occurring over a strong land-use gradi-ent encompassing intensively managed cropping systems andsemi-natural stands were investigated. Hence, the total num-ber of plots studied was 60 (5 transects× 12 ecosystems; Ta-ble 1 and Fig. 1). The cropping systems comprised multilayerand multi-crop agroforestry home gardens (Hom), monocul-ture coffee plantations (Cof) with dispersed shading trees,and maize fields (Mai) subject to regular albeit moderate fer-tilizer and pesticide applications. Plant litter is regularly re-moved from Cof and Mai sites. Home gardens are manuallyploughed, while combustion engine machinery is used forploughing coffee plantations and maize fields. Coffee plan-tations are irrigated with drip irrigation systems. Both Homand Cof sites still host indigenous forest trees that includeAlbizia schimperi, a species that may potentially fix atmo-spheric N. This is one of the five most abundant species in Ta
412 F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro
two and four of the Hom and Cof sites, respectively, makingup less than 25 % of the vegetation cover in all cases. Grass-lands (Gra) and savannas (Sav) are extensively managed bymeans of domestic grazing and occasional grass cutting,thus having significantly lower anthropogenic disturbancesthan cropping systems. Semi-natural ecosystems include sev-eral montane forest stands. These include lower montane(Flm), Ocotea (Foc), Podocarpus (Fpo), Erica (Fer), andalpine shrub vegetation Helichrysum (Hel). Even thoughlower montane forests are currently under protection they arestill subject to sporadic illegal logging. In addition to sam-pling undisturbed forest ecosystems of Ocotea and Podocar-pus, we purposely studied sites that had been affected by log-ging activities and fire events prior to the establishment of theKilimanjaro National Park (Soini, 2005): Ocotea (Fod) andPodocarpus (Fpd) (Table 1). Erica forests represent Africa’shighest forests in the subalpine zone. Higher above is thealpine zone, the realm of Helichrysum vegetation that is dom-inated by cushion plants and tussock grasses (Ensslin et al.,2015; Hemp, 2005). Potential ecosystem productivity anddecomposition rates show a hump-shaped pattern resemblingthat of precipitation (Supplement Fig. S1). It is interesting tosee the close match between the two variables along the el-evation range, although this trend weakens slightly towardshigher-elevation sites. Optimum growth and decompositionconditions are shown between 1800 and 2500 m a.s.l. Theselocations correspond to low-altitude forest ecosystems (Flmand Foc) that do not experience severe seasonal limitationsin moisture or temperature as is otherwise the case in lower-and higher-elevation systems that are moisture and tempera-ture limited, respectively (Becker and Kuzyakov, 2018).
Detailed physico-chemical characteristics of the dominantsoils are listed in Table 1. Soils in the Mount Kilimanjaroregion are mainly derived from volcanic rocks and ashes.The wide array of climatic conditions present along the el-evational gradient influence soil genesis, which results in theoccurrence of andosols at high elevations and soils of moreadvanced genesis at lower elevations (e.g. nitosols) (Majule,2003).
It is extremely difficult to provide reliable estimates ofboth fertilizer and pesticide rates used in small householdfarms in sub-Saharan Africa. This is because the actual useof these products is strongly dependent on both its availabil-ity in the local and/or regional market, the economic circum-stances of each individual farmer, and individual perceptionsabout their use (Saiz and Albrecht, 2016). The only sitesreceiving fertilizer are the two monocultures, maize (Mai)fields and coffee (Cof) plantations, and to a lesser extentthe home garden (Hom) sites. In the latter sites Gütlein etal. (2018) report that weed control is mainly done by hand,and the use of mineral or organic N fertilizers is low or non-existent. Extensively managed sites (i.e. Sav and Gra) receivevarying amounts of organic inputs as a result of grazing ac-tivities, but again, their actual rates are unknown. A more
Figure 1. Geographical distribution of investigated ecosystems:(a) along the elevational and land-use gradient. MAP denotesmean annual precipitation and MAT mean annual temperature.The colours of the boxes framing the ecosystem names match thecolours of symbols in the GeoTIFF panel below; (b) along thesouthern slope of Mount Kilimanjaro. Symbols represent individ-ual ecosystems (12) replicated five times (60 study sites in total).
detailed explanation on fertilizer and pesticide inputs used inthe region is provided in the Supplement (Table S1).
2.2 Sampling and analyses
Fieldwork took place in February and March in 2011 and2012. Sampling was conducted on 50× 50 m plots estab-lished at each of the 60 studied sites (12 ecosystems× 5 tran-sects). Surface litter and mineral topsoil (0–5 cm) were sam-pled at five locations (four corners and the central point) ateach plot. Additionally, fresh mature leaves of the five mostabundant plant species covering 80 % of total plant biomassper site were collected (Schellenberg Costa et al., 2017). Allsampled materials (leaves, litter, and soil) were air-dried un-til constant weight, and leaf material was subsequently oven-dried at 70 ◦C for 60 h prior to grinding. Soil was sieved to
F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro 413
2 mm with visible root fragments being further removed priorto grinding with a mixer mill (MM200, Retsch, Haan Ger-many). Soil pH was determined with a pH meter (MultiCalSenTix61, WTW, Weilheim, Germany) in a 0.01 M CaCl2 so-lution, with a CaCl2-to-soil ratio of 2 : 1. Particle size dis-tribution was determined gravimetrically using the pipettemethod (van Reeuwijk, 2002).
All soil, litter, and leaf samples were analysed with adry combustion elemental analyser (Costech InternationalS.p.A., Milano, Italy) fitted with a zero-blank autosam-pler coupled to a ThermoFinnigan DeltaPlus XL usingcontinuous-flow isotope ratio mass spectrometry (CF-IRMS)for the determination of the abundance of elemental C andN and their stable isotopic composition (δ13C, δ15N). Preci-sions (standard deviations) on internal standards for elemen-tal C and N concentrations and stable isotopic compositionswere better than 0.08 % and 0.2 ‰, respectively.
Natural 13C or 15N abundances are expressed in δ unitsaccording to Eq. (1):
δ(‰)= (Rsample−Rstandard/Rstandard)× 1000, (1)
where Rsample denotes the ratio 13C/12C or 15N/14N in thesample, and Rstandard denotes the ratios in Pee Dee Belem-nite or atmospheric N2 (international standards for C and N,respectively). The average values for the plant samples wereweighted considering their relative abundance at each site.Individual values for soil, litter, and leaves were averaged foreach plot.
In addition, both δ13C- and δ15N-based enrichment factors(ε) were calculated following Eqs. (2) and (3):
εC = δ13Clitter− δ
13Csoil, (2)
εN = δ15Nlitter− δ
15Nsoil. (3)
These were used as indicators for SOM decomposition dy-namics and ecosystem N status (Garten et al., 2008; Mariottiet al., 1981). Note that we use the stable isotopic values oflitter material rather than fresh leaves from various speciesto calculate enrichment factors, since litter provides a moreunbiased representation of the quality, quantity, and spatio-temporal dynamics of organic inputs entering the SOM pool(Saiz et al., 2015a).
2.3 Statistical analysis
The normal distribution of the data was confirmed withthe Shapiro–Wilk test. One-way ANOVA was performed totest for significant differences between ecosystems, whileTukey’s HSD was used as a post hoc procedure to test for sig-nificant differences across sites (P ≤ 0.05). Correlation anal-yses were performed to identify soil, foliar, and climatic vari-ables influencing soil δ15N values. Subsequently, a principalcomponent analysis (PCA) was conducted to reveal relation-ships between the main variables affecting soil δ15N values.The PCA was based on a correlation matrix including soil (C
and N concentrations, C/N ratio, δ13C, pH values, sand andclay contents) and climatic parameters (MAT and MAP). Astepwise multiple regression was used to identify the maindriving parameters determining soil δ15N across the eleva-tional transect. All statistical analyses were conducted withR (version 3.2.2; R Core Team, 2015).
3 Results
3.1 General soil characteristics
Soil C and N contents were the highest in forest ecosys-tems and showed a decreasing trend towards managed sites(i.e. home gardens, grasslands, coffee, and maize fields) (Ta-ble 1). Also, natural savannas and Helichrysum ecosystemshad lower soil C and N values compared to forest ecosys-tems. The low temperatures and sandy nature of the He-lichrysum sites play a strong role in their characteristicallylow productivity and moderate decomposition potentials (Ta-ble 1; Fig. S1), which unquestionably affects the compara-tively low soil C and N contents of these alpine systems.
An opposite trend to that of soil C and N abundance wasobserved for soil C/N ratios, whereby managed sites showedsignificantly lower values compared to those of semi-naturalecosystems. Soil pH values revealed acidic conditions at allsites, with the lowest values observed in forest sites havingcomparatively higher MAP (Table 1).
3.2 Variation of δ13C values along the elevational andland-use gradient
There were large variations in δ13C values along the ele-vational and land-use gradient, with distinct differences be-tween managed and semi-natural ecosystems (Fig. 2). Com-pared to soils and litter, leaves invariably showed the lowestδ13C values in all the studied ecosystems, with the exceptionof grasslands and savannas that exhibited lower soil δ13C val-ues than plant material.
The δ13C values of semi-natural ecosystems ranged be-tween−32.8 and−24.1 ‰ (mean±SE: soil−26.0±0.2 ‰;litter −27.2±0.2 ‰; leaves −29.3±0.3 ‰), showing a pro-gressive reduction with decreasing elevation (i.e. from 4500to 1750 m a.s.l.; Fig. S2). The variation in δ13C values wasmuch higher (−29.7 ‰ to −13.3 ‰) in managed ecosys-tems located at lower elevations (i.e. between 860 and1750 m a.s.l.; Fig. S2). The highest δ13C values were ob-served in C4-dominated ecosystems (i.e. savannas, maizefields, and grasslands; soil −16.8± 0.6 ‰, litter −19.3±0.8 ‰, leaves −18.8± 1.1 ‰), while lower δ13C valueswere obtained for coffee plantations and home gardens (soil−24.8±0.5 ‰, litter−27.2±0.4 ‰, leaves−27.3±0.4 ‰).Coffee plantations showed a slight influence of C4 vege-tation in the soil data as a result of grasses growing be-tween the rows of coffee plants. No significant variationswere observed between δ13C values of soils and those of
414 F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro
Figure 2. Variation in δ13C values for leaves, litter, and soil alongthe Kilimanjaro elevational and land-use gradient. Ecosystem datarepresent the average values of five sites (one per each transect),with each site being composed of five samples (n= 5). Box plotsshow median values per ecosystem with whiskers representing thefirst and third quartiles. Dots represent outliers. The shaded re-gion represents managed ecosystems (both intensively and exten-sively), while those unshaded indicate semi-natural ecosystems.Lower-case letters show significant differences between sampledmaterials within each ecosystem (one-way ANOVA followed byTukey’s HSD test as a post hoc procedure, P ≤ 0.05). The ecosys-tem acronyms used are as per Table 1. Mai, Cof, and Hom are man-aged cropping sites, Gra and Sav are extensively managed grass-lands and savannas, and the rest represent semi-natural ecosystems.Sites are ordered by increasing altitude.
litter and leaves in the ecosystems with a predominance ofC4 vegetation (savannas, maize fields, and grasslands). Ex-ploratory data analyses revealed that in most cases soil, litter,leaf, and climatic variables cross-correlated with each other(Table S2 in the Supplement).
Figure 3 shows relatively small variations in δ13C enrich-ment factors (> − 1.25 ‰) both in undisturbed semi-naturaland extensively managed sites along the elevational gradi-ent, while managed and disturbed sites show higher and morevariable δ13C enrichment factors.
3.3 Variation of δ15N values along the elevational andland-use gradient
Significantly higher δ15N values were observed for all sam-pled materials in the intensively managed (cropping) sys-tems compared to semi-natural and grass-dominated ecosys-tems (Fig. 4a). The δ15N values for managed systems rangedbetween −2.6 and 7.8 ‰ (mean±SE: soil 5.6± 0.3 ‰,litter 1.7± 0.5 ‰, leaves 2.0± 0.5 ‰). By contrast, semi-natural ecosystems had considerably lower δ15N values,which ranged from −5.0 ‰ to 3.6 ‰ (soil 1.5±0.2 ‰, litter
Figure 3. (a) Variation in δ13C-based enrichment factors(δ13Citter-soil) with elevation; (b) relationship between δ13C-basedenrichment factors (δ13Citter-soil) and SOC concentration (logSOC); and (c) relationship between δ13C-based enrichment factors(δ13Citter-soil) and soil C/N ratios. Note: a savanna site with largeC3 influence was removed from the figure for clarity.
F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro 415
Figure 4. Variation in δ15N values and δ15N-based enrichment factors along the Kilimanjaro elevational and land-use gradient. (a) Variationin δ15N values for leaves, litter, and soil material sampled along the Kilimanjaro elevational and land-use gradient. Box plots show medianvalues per ecosystem with whiskers representing the first and third quartiles. Dots represent outliers. Ecosystem data represent the averagevalues of five sites (one per each transect), with each site being composed of five samples. Lower-case letters show significant differencesbetween sampled materials within each ecosystem (one-way ANOVA followed by Tukey’s HSD test as a post hoc procedure, P ≤ 0.05).(b) Variation in δ15N-based enrichment factors (δ15Nlitter-soil) calculated for the different ecosystems along the elevational and land-use gra-dient. Dotted box plots indicate ecosystems dominated by C4 vegetation. Capital letters indicate significant differences between ecosystems(one-way ANOVA followed by Tukey’s HSD test as a post hoc procedure, P ≤ 0.05). The ecosystem acronyms used are the same as thosein Table 1. Sites are ordered by increasing altitude.
−2.1± 0.2 ‰, leaves −1.3± 0.3 ‰). Soil δ15N values weresignificantly higher than those of leaves and litter across allthe ecosystems studied, with the only exception being agro-forestry home gardens (Fig. 4a). δ15N values of leaves andlitter did not show significant differences within any givenecosystem.
Calculated δ15N-based enrichment factors showed highvariability across all ecosystems with values ranging from−7.5 ‰ to−1.6 ‰ (Fig. 4b). A differentiation between man-aged and natural ecosystems was less clear than for δ15N val-ues. The most negative enrichment factors (< −4.0 ‰) wereobserved for Helichrysum, Erica, Podocarpus disturbed,and grass-dominated ecosystems (savannas and grasslands).These enrichment factors were significantly less negative formontane forests at lower elevations (Podocarpus, Ocotea,and lower montane) and intensively managed (cropping) sys-tems (i.e. home garden, coffee, and maize; Fig. 4b).
3.4 Impacts of soil and climatic variables on soil δ15Nvalues
Two principal components (PCs) explained 78.3 % of thetotal soil δ15N variation (Fig. 5). The first component ex-plained 55.8 % of the variability, and included soil chem-istry and climatic variables (soil C and N concentrations,soil C/N ratio, soil pH, soil δ13C, MAP, and MAT). Highlysignificant correlations (P <0.001) were obtained betweenPC 1 and the above factors (r = 0.93, 0.93, 0.61, −0.87,−0.76, 0.87, and −0.63, respectively; Table S3). The sec-
ond component explained an additional 22.5 % of soil δ15Nvariability and included soil texture (clay and sand con-tents) and MAT. These variables were highly correlated withPC 2 (r =−0.84, 0.82, and −0.65; Table S3). The princi-pal component bi-plot showed a strong grouping betweenmanaged and semi-natural ecosystems (Fig. 5). Managedsites clustered around MAT, soil δ13C, and soil pH, whileC4-dominated ecosystems (grassland, savannas, and maizefields) were preferentially influenced by the latter two vari-ables. In contrast, semi-natural montane forest ecosystemsrather grouped around soil chemical properties such as C andN contents, C/N ratio, and MAP, while alpine Helichrysumecosystems clustered around soil sand content.
In addition to PCA, multiple regression analyses were per-formed using a stepwise procedure that identified soil N con-tent and MAT as the main driving variables explaining thevariation in soil δ15N. A paraboloid model explained 68 %of this variability (P <0.05; Fig. 6). The combination ofrelatively high soil N contents (1 % to 3 %) and low MAT(up to 14 ◦C) invariably corresponded to low soil δ15N val-ues (<2 ‰) characteristic of semi-natural ecosystems. Con-versely, the relatively high soil δ15N values (>2 ‰) observedin managed ecosystems corresponded to low soil N contents(<1 %) and comparatively high MAT (17 to 25 ◦C).
The relationship between soil δ15N values and climaticand edaphic variables provided valuable information aboutpotentially different SOM dynamics in the various ecosys-tems studied, with data showing a clear differentiation be-tween semi-natural and managed ecosystems (Fig. S4). The
416 F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro
Figure 5. Principal component analysis bi-plot for soil and climatevariables potentially controlling soil δ15N. Symbols are as per allprevious figures. Acronyms are as per Table 1. C/N: soil C/N ratio,C: soil carbon content, N: soil nitrogen content, MAP: mean annualprecipitation, clay: soil clay content, MAT: mean annual tempera-ture, δ13C: soil δ13C, and pH: soil pH.
former is characterized by comparatively higher C/N ratiosand lower δ15N values (averaging 15.5 ‰ and 1.5 ‰, re-spectively), while the latter showed lower C/N ratios andhigher soil δ15N values (averaging 11.9 ‰ and 3.5 ‰, re-spectively). Managed ecosystems further grouped into inten-sively cropped (home gardens, maize fields, and coffee plan-tations) and extensively managed grass-dominated ecosys-tems (savannas and grasslands).
4 Discussion
4.1 Factors influencing the variation of δ13C valuesalong the elevational and land-use gradient
The δ13C values of leaves in C3-dominated (semi-natural)ecosystems at Mount Kilimanjaro increased with elevation(Figs. 1 and S2), which is in agreement with findings fromother mountainous ecosystems in the tropics, Europe, andNorth America (Bird et al., 1994; Körner et al., 1991; Ortizet al., 2016; Zhou et al., 2011; Zhu et al., 2009). The widerscatter of δ13C values observed in leaves relative to soils ismost certainly due to the inherently large (interspecific andintraspecific) variability of δ13C in plants (Bird et al., 1994).Different tissues within the plant can present widely diver-gent δ13C values as a result of fractionation processes as-sociated with the C compounds involved in their construc-
Figure 6. Measured and modelled soil δ15N values predicted asa function of soil N abundance and mean annual temperature(MAT). Data points are classified by generic land uses (i.e. in-tensively managed cropping sites, extensively managed grasslandand savannas, and semi-natural ecosystems) observed along the el-evational and land-use gradient. The regression takes the follow-ing form: soil δ15N= 1.10+ 0.49 (MAT)− 1.86 (soil N)− 0.01(MAT)2
+ 0.14 (soil N)2; (r2 adj= 0.68, P <0.05, n= 60).
tion (Dawson et al., 2002). Moreover, other factors includinglight intensity, humidity, and the reutilization of previouslyrespired low 13C-CO2 within the canopy may further con-tribute to the variability of δ13C in leaf tissues (Ometto et al.,2006; van der Merwe and Medina, 1989).
While fractionation effects preclude a straightforward in-terpretation of δ13C of SOM, this variable provides an inte-grated measure of the isotopic composition of the precursorbiomass at the ecosystem level (Bird et al., 2004; Saiz et al.,2015a). Mass balance calculations that assume (i) 5 % (w/w)average root mass (<2 mm) in soil samples and (ii) leaveshaving similar isotopic signals as roots show that the removalof visible sieved roots might cause a very small effect onsoil isotopic values. This would amount to values ∼ 0.15 ‰higher than the original soil isotopic values, with such dis-crepancy being even smaller if root samples were consid-ered to have values 0.5 ‰–1 ‰ higher than leaves, as is com-monly reported in the literature (calculations not shown). Be-sides the natural variability of soil δ13C values observed inC3-dominated semi-natural ecosystems, there were distinctpatterns in the δ13C values of soil samples collected in exten-sively managed, low-elevation ecosystems where woody andgrass vegetation coexist (i.e. grasslands and savannas), whichindicates the strong influence exerted by C4 vegetation on theC isotopic composition of all sampled materials (Fig. 2). Theresults obtained in semi-natural ecosystems at Mount Kili-manjaro fit well within the interpretative framework for ele-
F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro 417
vational soil δ13C data proposed by Bird et al. (1994). Theseauthors suggest that besides temperature and atmosphericpressure, other primary factors influencing soil δ13C valuesare the age and degree of decomposition of SOM and vari-ables related to the characteristics of the canopy, includingthe proportion of respired CO2 that is recycled during photo-synthesis, the relative contribution of leaf and woody litter toSOM, and soil moisture.
Besides the factors explained above, soil δ13C values arestrongly influenced by the balance between ecosystem C in-puts and outputs. It seems reasonable to assume that in thecase of natural ecosystems there may be a steady state be-tween SOM inputs and decomposition rates. This should bein contrast with the typically altered nutrient dynamics of dis-turbed systems, particularly those under agricultural manage-ment (Wang et al., 2018). Low fractionation factors in δ13Care commonly reported between plant material and topsoilsin natural systems, mainly because of the relatively limitedhumification of recent organic matter prevalent in topsoils(Acton et al., 2013; Wang et al., 2018). Thus, we hypoth-esized that if C inputs and outputs were roughly in balance,then the difference in δ13C values between plant material andtopsoil would be smaller in undisturbed sites compared tomanaged or disturbed sites. The results shown in Fig. 3 agreewell with this notion.
Soil δ13C values decreased with increasing MAP and de-creasing MAT, which also corresponded with higher SOCcontents (Fig. S3). This suggests that the relatively coolerand wetter conditions of high-elevation semi-natural for-est ecosystems (i.e. Foc, Fpo) promote the accumulation ofSOM, which is similar to previous findings of work con-ducted along elevational gradients (Bird et al., 1994; Kohn,2010). Compared to high-elevation locations, the climaticconditions of mid-elevation ecosystems are more favourablefor the activities of SOM decomposers, as these sites are con-sistently warmer and drier than the characteristically cool andoccasionally waterlogged high-altitude ecosystems (Fig. S1;Becker and Kuzyakov, 2018; Borken and Matzner, 2009;Garten et al., 2009; Kirschbaum, 1995; Leirós et al., 1999).The comparatively high soil δ13C values observed in the dis-turbed Podocarpus (Fpd) and Erica forest (Fer) plots mayhave been partly caused by recurrent fire events (Hemp,2005) leading to reduced SOC contents and higher C/N ra-tios (Saiz et al., 2015a). Further variations in soil δ13C val-ues could also be related to the biochemical composition ofthe precursor biomass. For instance, herbaceous vegetationis pervasive at high elevations and contains relatively lowamounts of lignin – an organic compound characteristicallydepleted in 13C (Benner et al., 1987). This may contributeto explaining the higher δ13C values observed in plant andsoil materials in alpine ecosystems dominated by Helichry-sum vegetation compared to forest ecosystems at lower ele-vations (Fig. 2).
Elevation also has a strong influence on the seasonal lit-terfall dynamics observed in Mount Kilimanjaro and thus
may have significant implications in the SOM cycling acrossthe various ecosystems. Becker et al. (2015) suggest that thelarge accumulation of particulate organic matter observed atthe end of the dry season in low- and mid-altitude ecosystemsmay result in the increased mineralization of easily avail-able substrates (Mganga and Kuzyakov, 2014) and nutrientleaching (Gütlein et al., 2018) during the following wet sea-son. Agricultural practices such as the removal of biomassor ploughing deplete SOM, particularly in intensively man-aged systems (i.e. maize, home gardens, and coffee plan-tations), thus leading to lower SOC contents and C/N ra-tios and slightly higher soil δ13C values than those observedin semi-natural ecosystems at comparable elevations (e.g.lower montane forests; Fig. S3). Indeed, the relationship be-tween δ13C enrichment factors and soil C/N ratios shownin Fig. 3 is quite informative regarding SOM dynamics. Aspreviously mentioned, soil C/N ratios provide a good in-dication of SOM decomposition processes, typically show-ing comparatively low values in managed and disturbed sys-tems. These correspond well with sites having large enrich-ment factors (< − 1.25 ‰; i.e. intensively managed and dis-turbed sites), which agrees with the notion of altered SOMdynamics. Therefore, besides the systematic removal of plantbiomass characteristic of agricultural systems, annual litter-fall patterns may also explain the comparatively lower con-tents of C and N observed in the topsoils of intensively man-aged sites (Table 1; Figs. S3, S4). Moreover, low-elevationecosystems contain a variable mixture of C3 and C4 vege-tation, which have been shown to have differential mineral-ization dynamics as demonstrated by incubation experiments(Wynn and Bird, 2007) and field-based research (Saiz et al.,2015a).
Our data show strong relationships between temperatureand variables directly related to SOM dynamics such as soilδ13C, C, N, and C/N ratios (Table S2). These results agreewell with recent findings by Becker and Kuzyakov (2018),who studied SOM decomposition dynamics at these verysites. An important finding revealed by that study is seasonalvariation in temperature being a major factor controlling lit-ter decomposition. Their study shows that small seasonalvariations in temperature observed at high-elevation sites ex-ert a strong effect on litter decomposition rates. Therefore,the authors argue that the projected increase in surface tem-perature may result in potentially large soil C losses at thesesites due to the comparatively strong temperature sensitivityto decomposition that is commonly observed at low tempera-tures and at high-elevation sites (Blagodatskaya et al., 2016).
Savannas and grasslands are subject to recurrent fireevents, and thus the soils of these ecosystems may poten-tially contain significant amounts of fire-derived (pyrogenic)C (Saiz et al., 2015b). This can be partly demonstrated by thehigher soil C/N ratios observed in these ecosystems com-pared to C4-dominated agricultural systems protected fromfire (e.g. maize plantations; Fig. S3d). Moreover, the δ13Cvalues of soils in grasslands and savannas were lower than
418 F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro
those of leaves, which may be due to the savanna isotopedisequilibrium effect (SIDE) (Bird and Pousai, 1997; Saiz etal., 2015b). The latter concept explains the difference in Cisotopic composition between the precursor vegetation andpyrogenic C compounds produced during the combustion ofbiomass. Saiz et al. (2015b) have demonstrated that savannafires produce pyrogenic C that is relatively 13C depleted withrespect to the precursor biomass. Furthermore, the combus-tion of C4 vegetation produces finer pyrogenic C particlesthan woody biomass, resulting in the preferential export ofgrass-derived pyrogenic particles from the site of burning,which further enhances the depletion of 13C in these soils(Saiz et al., 2018).
4.2 Variation of δ15N values along the elevational andland-use gradient
The δ15N values of leaves, litter, and topsoil presented here(Fig. 4a) agree well with the range of data reported from ear-lier investigations in the same study region (Amundson et al.,2003; Zech et al., 2011), but with our study involving moreecosystems, replicate sites, and a far larger spatial samplingdomain. Overall, the δ15N values for montane tropical for-est ecosystems in Mount Kilimanjaro are considerably lowerthan the mean values reported for a broad variety of tropi-cal lowland forests worldwide (soil values ranging from 3 ‰to 14 ‰; de Freitas et al., 2015; Martinelli et al., 1999; Nar-doto et al., 2014; Piccolo et al., 1996; Sotta et al., 2008).Rather, the δ15N values observed in the montane forests in-vestigated are in the same range of temperate forest ecosys-tems reported in a comprehensive literature review by Mar-tinelli et al. (1999). These authors argue that, compared totropical lowland forests, the lower δ15N values of temper-ate and montane tropical forests result from their lower Navailability and thus lower ecosystem N losses. However, thishypothesis may not completely hold for the montane for-est ecosystems of our study, since Gütlein et al. (2018) re-ported elevated soil NO−3 and DON concentrations at deepsoil solution (80 cm) and significant nitrogen leaching ratesof 10–15 kg N ha−1 yr−1. The relatively low δ15N-based en-richment factors observed in the lower montane, Ocotea, andundisturbed Podocarpus forest (Fig. 4b) were probably dueto the prevalence of biological dinitrogen fixation (BNF) atthese ecosystems. The assumption of significant BNF is sup-ported by leaf δ15N values close to 0 ‰ (Fig. 4a) and is inline with previous works (Craine et al., 2015a; Nardoto etal., 2014; Robinson, 2001). Furthermore, sporadic measure-ments of N compounds in rainfall and throughfall conductedat our forest sites showed substantial input of N via atmo-spheric deposition, which may be of the order of N leach-ing losses (unpublished results). This agrees well with find-ings from Bauters et al. (2018) reporting 18 kg N ha−1 yr−1 Ninputs via wet deposition into tropical forests of the CongoBasin, which are predominantly derived from biomass burn-ing and long-range atmospheric transport. High N inputs
into these forest ecosystems are likely to be in a similarrange as N outputs (prevailed by leaching losses, particu-larly where MAP is highest; Gütlein et al., 2018), and there-fore they would not translate to strong effects on ecosystemδ15N values. The significantly more negative enrichment fac-tors observed in the disturbed Podocarpus and Erica forests(Fig. 4b) may be related to past fire events (Hemp, 2005;Zech et al., 2011). Burning of vegetation may cause lossesof 15N-depleted NOχ gas and N leachate, resulting in highersoil δ15N values, thus producing variations in δ15N-based en-richment factors (Zech et al., 2011).
Previous studies have shown that δ15N values generally in-crease with land-use intensification (Martinelli et al., 1999;Stevenson et al., 2010), which corresponds well with themore positive δ15N values observed in the intensively man-aged agricultural systems occurring at the mountain’s footslope (Fig. 4a). Indeed, agronomic practices such as fertiliza-tion, the removal of plant material after harvest, and plough-ing are factors known to impact N turnover processes thatstrongly affect δ15N values (Bedard-Haughn et al., 2003;Saiz et al., 2016). However, our values are in the lower rangeof published data for other land-use gradients (Aranibar etal., 2008; Eshetu and Högberg, 2000; Traoré et al., 2015)and may partly be the result of comparably low to moderateorganic and inorganic N fertilization rates currently appliedin the region (anecdotal evidence gathered by the authors andfound in the Supplement). Additionally, the nitrogen isotopicsignal of mineral fertilizers commonly used in the region is∼ 0 ‰ (Bateman and Kelly, 2007), and thus it may not ex-ert a significant additional bias on the interpretation of soilδ15N values. However, the addition of manure (δ15N∼ 8 ‰)in Hom systems, although used in low quantities (Gütlein etal., 2018), may well have contributed to the high δ15N val-ues observed in this ecosystem (Fig. 4). Also, we suggest thatthe use of pesticides may not pose a strong bias in our iso-topic results since their use is limited to intensively managedsites, and the actual isotopic values of pesticides work in theopposite direction to the observed data (Fig. 4; Supplement).
Compared to other low-elevation managed stands such ashome gardens and coffee plantations, the higher δ15N-basedenrichment factors observed in maize fields and in grass-dominated ecosystems (grasslands and savannas) (Fig. 4b)may be related to both the organic inputs resulting from graz-ing activities and the influence of C4 vegetation. Both Arani-bar et al. (2008) and Wang et al. (2010) have suggested thatvariations in δ15N values within a given ecosystem could bedue to C3 and C4 plants preferentially absorbing chemicalforms of N with differing 15N abundances. Moreover, recur-rent fires characteristic of tropical grasslands and savannasmay have also influenced their comparatively high soil δ15N,causing the relatively high δ15N-based enrichment factors.
F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro 419
4.3 Factors controlling soil δ15N along the elevationaland land-use gradient
The strong controlling effects exerted by climatic andedaphic factors on soil δ15N values agree well with numerousprevious works (Amundson et al., 2003; Conen et al., 2013;Eshetu and Högberg, 2000; Martinelli et al., 1999; Steven-son et al., 2010). The principal component analysis of factorscontrolling soil δ15N revealed a strong clustering betweenmanaged and semi-natural ecosystems (Fig. 5), which wasalso reflected in the multiple regression analysis and graph-ical representation depicting soil δ15N as a function of soilN concentration and MAT (Fig. 6). Semi-natural ecosystemswere characterized by relatively low soil δ15N values and oc-curred across a broad range of soil N contents in locationswith low to medium MAT. By contrast, intensively managedecosystems had higher soil δ15N values and corresponded tolocations with low soil N contents and high MAT. The nega-tive correlation of δ15N values with soil nitrogen content andthe positive correlation with mean annual temperature sug-gest reduced mineralization rates and thus limited nitrogenavailability, at least in high-elevation ecosystems.
The sharp contrast observed both in soil C/N ratios andδ15N values between managed and semi-natural ecosystemsoffers additional useful information about their potentiallycontrasting SOM dynamics (Fig. S4d). Intensively managedsites consistently showed low soil C/N ratios and high soilδ15N values, which may initially suggest a more open Ncycle and potentially greater N losses as reported by Ger-schlauer et al. (2016) for some of these ecosystems. Thismay be due to the C limitation of heterotrophic microbial Nretention under low C/N ratios (Butterbach-Bahl and Dan-nenmann, 2012). However, nitrate leaching is quite a rel-evant process that discriminates only slightly against 15N(Denk et al., 2017), which may confound the interpretationof soil δ15N values. Indeed, Gütlein et al. (2018) have re-cently shown that nitrate leaching may be quite significant inMount Kilimanjaro’s semi-natural forests. Therefore, at leastin these ecosystems, claims about the nature of the N cycle(i.e. open or closed) should not be made solely on the basisof soil δ15N.
Grass-dominated ecosystems (grasslands and savannas)were noticeably different to the intensively managed crop-lands, as demonstrated by the higher soil C/N ratios andlower soil δ15N of the former, which suggests a lower degreeof decomposition of organic matter and potentially lowerN turnover rates (Saiz et al., 2016). Within the intensivelymanaged sites, the stands under maize cultivation show aninteresting case of enhanced SOM dynamics. These sitesare under an intensive management regime that involves theremoval of aboveground vegetation after harvest. This factcombined with the faster decomposition rates reported forC4-derived SOM (Saiz et al., 2015a, 2016; Wynn and Bird,2007) may invariably lead to their characteristically low SOCand N contents (Table 1; Figs. S3, S4). Furthermore, low soil
C/N ratios have been reported to enhance gaseous losses insemi-arid systems, which leads to increased soil δ15N values(Aranibar et al., 2004) and may explain why maize standsshowed the highest soil δ15N values of all the land uses stud-ied.
Semi-natural ecosystems showed rather high soil C/N ra-tios and low soil δ15N values compared to managed sites(Fig. S4d). The more humid and cooler conditions preva-lent in forest ecosystems may limit decomposition processes,thereby contributing significantly to their higher SOM abun-dance (Table 1). A small variation range in soil δ15N val-ues was also reported by Zech et al. (2011) for semi-naturalecosystems (Foc and Fpo) when working along the sameland-use and elevation gradient. Like us, these authors alsoobserved a strong significant correlation of soil δ15N withMAT, but not with MAP (Table S2). Additionally, site-specific soil characteristics and the structural compositionof vegetation have a strong influence on ecosystem nutri-ent dynamics (Saiz et al., 2012, 2015a). Ecosystem distur-bances (e.g. fire, selective logging, etc.) cause changes invegetation cover that affect SOM cycling and may translateinto variations in soil C/N ratios (Saiz et al., 2016). BothOcotea and Podocarpus forests contain disturbed (Fod, Fpd)and undisturbed stands (Foc, Fpo), though only the Podocar-pus ecosystems allow for a general overview of disturbanceimpacts on SOM-related properties. While changes in theisotopic composition of C and N were not significant, soilC/N ratios were heavily influenced by disturbance (Fig. S4).Compared to non-disturbed sites, the lower C and N con-tents observed in the soil of disturbed ecosystems indicate re-duced OM inputs to the soil and/or enhanced decompositionof SOM (Table 1). The higher soil C/N ratios observed in thePodocarpus disturbed and Erica forests may well be the re-sult of fire, which may preferentially promote N losses whileaccruing relatively recalcitrant C forms (i.e. pyrogenic C).Woody biomass combustion produces pyrogenic C that ac-cumulates preferentially close to the site of production (Saizet al., 2018), thus likely contributing to the higher soil C/Nratios observed at these disturbed ecosystems. The lowestsoil C/N ratios among all semi-natural ecosystems were ob-served at the alpine Helichrysum sites, which may relate totheir characteristically sparse vegetation and extremely lowMAT. Under such circumstances, soil development, biomassinputs, decomposition processes, and thus soil N turnovermay be strongly limited, as was confirmed by a recent studyconducted at one of these sites (Gütlein et al., 2017).
5 Conclusions
The variations in δ13C and δ15N values combined with theinterpretation of other indices, such as δ13C- and δ15N-basedenrichment factors and soil C/N ratios, enabled a qualitativecharacterization of regional differences in C and N dynam-
420 F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro
ics as affected by vegetation characteristics, environmentalconditions, and management activities.
Our data show that SOM contents are higher in cold andwet high-elevation ecosystems than at low-elevation man-aged sites. Management practices such as tillage, harvest,and vegetation burning promote the loss of OM, with SOMdecomposition being further enhanced by the warm and mod-erately wet conditions of the mountain’s foot slope. Based onour results, we suggest that besides management, increasingtemperatures in a changing climate may promote C and Nlosses, thus altering the otherwise stable SOM dynamics ofMount Kilimanjaro’s forest ecosystems. Moreover, the cur-rent situation of low N inputs in managed systems of sub-Saharan Africa is likely to change, since national efforts thataim to increase fertilizer use are currently at <10 % of rec-ommended rates (Hickman et al., 2014). Therefore, our datamay also be valuable as a generic reference for low-elevationtropical agrosystems managed under low N inputs, whileit may also allow for the monitoring of expected changesin agricultural management and associated impacts on theecosystem N cycle through the study of variation in δ15Nvalues.
In addition to climatic and edaphic factors, δ15N values ofplant and soil material can largely depend on both the amountand δ15N signal of atmospheric deposition and BNF, whichhighlights the importance of conducting additional measure-ments of site-specific N cycling when comparing ecosystemδ15N values across different biomes and regions. The com-bination of qualitative isotope natural abundance studies at alarge number of sites (this study) with more elaborated quan-titative process studies using enriched isotope labelling andN losses on a lower number of selected sites represents anideal approach to characterize the ecosystem C and N cy-cling of the larger Mount Kilimanjaro region with its diverseecosystems, climate, and management.
Data availability. Research data can be requested and accessedthrough the following link: https://www.kilimanjaro.biozentrum.uni-wuerzburg.de/Data/Data.aspx (Gerschlauer and Kiese, 2016).
Supplement. The supplement related to this article is availableonline at: https://doi.org/10.5194/bg-16-409-2019-supplement.
Author contributions. FG contributed to design, performed thestudy, and co-wrote the paper; GS contributed to analyses and co-wrote the paper; DSC and MK provided plant samples and con-tributed to writing; MD contributed to writing; and RK designedthe study and contributed to analyses and writing.
Competing interests. The authors declare that they have no conflictof interest.
Acknowledgements. This study was funded by the GermanResearch Foundation (DFG: KI 1431/1-1 and KI 1431/1-2)within research unit 1246 (KiLi) and supported by the TanzanianCommission for Science and Technology (COSTECH), theTanzania Wildlife Research Institute (TAWIRI), and the MountKilimanjaro National Park (KINAPA). In addition, the authorsthank Andreas Hemp for the selection and preparation of theresearch plots, Bernd Huwe for the correction of soil texture data,all local helpers in Tanzania, and the assistants in the laboratoryof IMK-IFU in Germany. Technical support by the Center ofStable Isotopes at KIT/IMK-IFU is gratefully acknowledged.Further thanks go to the following persons from the KiLi project:Tim Appelhans and Thomas Nauss, Jie Zhang, Gemma Rutten,and Andreas Hemp for providing georeferenced points under-lying the GeoTIFF in Fig. 1b. We also thank four anonymousreviewers and Jonathan Wynn for insightful comments on the paper.
The article processing charges for this open-accesspublication were covered by a ResearchCentre of the Helmholtz Association.
Edited by: Yakov KuzyakovReviewed by: four anonymous referees
References
Abaker, W. E., Berninger, F., Saiz, G., Braojos, V., and Starr, M.:Contribution of Acacia senegal to biomass and soil carbon inplantations of varying age in Sudan, Forest Ecol. Manag., 368,71–80, https://doi.org/10.1016/j.foreco.2016.03.003, 2016.
Abaker, W. E., Berninger, F., Saiz, G., Pumpanen, J., and Starr, M.:Linkages between soil carbon, soil fertility and nitrogen fixationin Acacia senegal plantations of varying age in Sudan, PeerJ, 6,e5232, https://doi.org/10.7717/peerj.5232, 2018.
Acton, P., Fox, J., Campbell, E., Rowe, H., and Wilkinson, M.: Car-bon isotopes for estimating soil decomposition and physical mix-ing in well-drained forest soils, J. Geophys. Res.-Biogeo., 118,1532–1545, 2013.
Amundson, R., Austin, A. T., Schuur, E. A. G., Yoo, K.,Matzek, V., Kendall, C., Uebersax, A., Brenner, D., and Bais-den, W. T.: Global patterns of the isotopic composition ofsoil and plant nitrogen, Global Biogeochem. Cy., 17, 1031,https://doi.org/10.1029/2002GB001903, 2003.
Appelhans, T., Mwangomo, E., Otte, I., Detsch, F., Nauss, T., andHemp, A.: Eco-meteorological characteristics of the southernslopes of Kilimanjaro, Tanzania, Int. J. Climatol., 36, 3245–3258, https://doi.org/10.1002/joc.4552, 2016.
Aranibar, J. N., Otter, L., Macko, S. A., Feral, C. J., Epstein, H.E., Dowty, P. R., Eckardt, F., Shugart, H. H., and Swap, R. J.:Nitrogen cycling in the soil–plant system along a precipitationgradient in the Kalahari sands, Glob. Change Biol., 10, 359-373.https://doi.org/10.1111/j.1365-2486.2003.00698.x, 2004.
Aranibar, J. N., Anderson, I. C., Epstein, H. E., Feral, C. J. W.,Swap, R. J., Ramontsho, J., and Macko, S. A.: Nitrogen iso-tope composition of soils, C3 and C4 plants along land usegradients in southern Africa, J. Arid Environ., 72, 326–337,https://doi.org/10.1016/j.jaridenv.2007.06.007, 2008.
F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro 421
Bai, E., Houlton, B. Z., and Wang, Y. P.: Isotopic identificationof nitrogen hotspots across natural terrestrial ecosystems, Bio-geosciences, 9, 3287–3304, https://doi.org/10.5194/bg-9-3287-2012, 2012.
Bateman, A. S. and Kelly, S. D.: Fertilizer nitrogen isotope signa-tures, Isot. Environ. Heal. S., 43, 237–247, 2007.
Bauters, M., Drake, T. W., Verbeeck, H., Bodé, S., Hervé-Fernández, P., Zito, P., Podgorski, D. C., Boyemba, F.,Makelele, I., Ntaboba, L. C., Spencer, R. G. M., andBoeckx, P.: High fire-derived nitrogen deposition on cen-tral African forests, P. Natl. Acad. Sci. USA, 115, 549–554,https://doi.org/10.1073/pnas.1714597115, 2018.
Becker, J., Pabst, H., Mnyonga, J., and Kuzyakov, Y.: Annual lit-terfall dynamics and nutrient deposition depending on elevationand land use at Mt. Kilimanjaro, Biogeosciences, 12, 5635–5646,https://doi.org/10.5194/bg-12-5635-2015, 2015.
Becker, J. N. and Kuzyakov, Y.: Teatime on Mount Kilimanjaro:Assessing climate and land-use effects on litter decompositionand stabilization using the Tea Bag Index, Land Degrad. Dev.,29, 2321–2329, 2018.
Bedard-Haughn, A., van Groenigen, J. W., and van Kessel, C.:Tracing 15N through landscapes: potential uses and precau-tions, J. Hydrol., 272, 175–190, https://doi.org/10.1016/S0022-1694(02)00263-9, 2003.
Benner, R., Fogel, M. L., Sprague, E. K., and Hodson,R. E.: Depletion of 13C in lignin and its implicationsfor stable carbon isotope studies, Nature, 329, 708–710,https://doi.org/10.1038/329708a0, 1987.
Bird, M. I., Haberle, S. G., and Chivas, A. R.: Effect of altitudeon the carbon-isotope composition of forest and grassland soilsfrom Papua New Guinea, Global Biogeochem. Cy., 8, 13–22,https://doi.org/10.1029/93GB03487, 1994.
Bird, M. I. and Pousai, P.: Variations of δ13C in the surface soilorganic carbon pool, Global Biogeochem. Cy., 11, 313–322,https://doi.org/10.1029/97GB01197, 1997.
Bird, M. I., Veenendaal, E. M., and Lloyd, J. J.: Soil carbon in-ventories and δ13C along a moisture gradient in Botswana,Glob. Change Biol., 10, 342–349, https://doi.org/10.1046/j.1365-2486.2003.00695.x, 2004.
Blagodatskaya, E., Blagodatsky, S., Khomyakov, N., Myachina, O.,and Kuzyakov, Y.: Temperature sensitivity and enzymatic mech-anisms of soil organic matter decomposition along an altitudi-nal gradient on Mount Kilimanjaro, Sci. Rep.-UK, 6, 22240,https://doi.org/10.1038/srep22240, 2016.
Booth, M. S., Stark, J. M., and Rastetter, E.: Controls onNitrogen Cycling in Terrestrial Ecosystems: A SyntheticAnalysis of Literature Data, Ecol. Monogr., 75, 139–157,https://doi.org/10.1890/04-0988, 2005.
Borken, W. and Matzner, E.: Reappraisal of drying and wet-ting effects on C and N mineralization and fluxes in soils,Glob. Change Biol., 15, 808–824, https://doi.org/10.1111/j.1365-2486.2008.01681.x, 2009.
Butterbach-Bahl, K. and Dannenmann, M.: Soil Carbon and Ni-trogen Interactions and Biosphere-Atmosphere Exchange of Ni-trous Oxide and Methane, in: Recarbonization of the Biosphere,edited by: Lal, R., Lorenz, K., Hüttl, R. F., Schneider, B. U., andvon Brau, J., Springer, the Netherlands, 429–443, 2012.
Cernusak, L. A., Ubierna, N., Winter, K., Holtum, J. A.M., Marshall, J. D., and Farquhar, G. D.: Environmen-
tal and physiological determinants of carbon isotope dis-crimination in terrestrial plants, New Phytol., 200, 950–965,https://doi.org/10.1111/nph.12423, 2013.
Conen, F., Yakutin, M. V., Carle, N., and Alewell, C.: δ15N nat-ural abundance may directly disclose perturbed soil when re-lated to C : N ratio, Rapid Commun. Mass Sp., 27, 1101–1104,https://doi.org/10.1002/rcm.6552, 2013.
Craine, J. M., Brookshire, E. N. J., Cramer, M. D., Hasselquist, N.J., Koba, K., Marin-Spiotta, E., and Wang, L.: Ecological inter-pretations of nitrogen isotope ratios of terrestrial plants and soils,Plant Soil, 396, 1–26, https://doi.org/10.1007/s11104-015-2542-1, 2015a.
Craine, J. M., Elmore, A. J., Wang, L., Augusto, L., Baisden, W. T.,Brookshire, E. N. J., Cramer, M. D., Hasselquist, N. J., Hobbie,E. A., Kahmen, A., Koba, K., Kranabetter, J. M., Mack, M. C.,Marin-Spiotta, E., Mayor, J. R., McLauchlan, K. K., Michelsen,A., Nardoto, G. B., Oliveira, R. S., Perakis, S. S., Peri, P. L.,Quesada, C. A., Richter, A., Schipper, L. A., Stevenson, B. A.,Turner, B. L., Viani, R. A. G., Wanek, W., and Zeller, B.: Con-vergence of soil nitrogen isotopes across global climate gradi-ents, Sci. Rep.-UK, 5, 8280, https://doi.org/10.1038/srep08280,2015b.
Dawson, T. E., Mambelli, S., Plamboeck, A. H., Tem-pler, P. H., and Tu, K. P.: Stable Isotopes in PlantEcology, Annu. Rev. Ecol. Syst., 33, 507–559,https://doi.org/10.1146/annurev.ecolsys.33.020602.095451,2002.
de Freitas, A. D. S., de Sá Barretto Sampaio, E. V., de SouzaRamos, A. P., de Vasconcellos Barbosa, M. R., Lyra, R. P., andAraújo, E. L.: Nitrogen isotopic patterns in tropical forests alonga rainfall gradient in Northeast Brazil, Plant Soil, 391, 109–122,https://doi.org/10.1007/s11104-015-2417-5, 2015.
Denk, T. R. A., Mohn, J., Decock, C., Lewicka-Szczebak, D.,Harris, E., Butterbach-Bahl, K., Kiese, R., and Wolf, B.:The nitrogen cycle: A review of isotope effects and iso-tope modeling approaches, Soil Biol. Biochem., 105, 121–137,https://doi.org/10.1016/j.soilbio.2016.11.015, 2017.
Duane, W., Pepin, N., Losleben, M., and Hardy, D.: General Char-acteristics of Temperature and Humidity Variability on Kiliman-jaro, Tanzania, Arct. Antarct. Alp. Res., 40, 323–334, 2008.
Ensslin, A., Rutten, G., Pommer, U., Zimmermann, R., Hemp, A.,and Fischer, M.: Effects of elevation and land use on the biomassof trees, shrubs and herbs at Mount Kilimanjaro, Ecosphere, 6,1–15, https://doi.org/10.1890/ES14-00492.1, 2015.
Eshetu, Z. and Högberg, P.: Effects of land use on 15N natural abun-dance of soils in Ethiopian highlands, Plant Soil, 222, 109–117,2000.
FAO and JRC: Global forest land-use change 1990–2005, in: FAOForestry Paper No. 16, edited by: Lindquist, E. J., D’Annunzio,R., Gerrand, A., MacDicken, K., Achard, F., Beuchle, R., Brink,A., Eva, H. D., Mayaux, P., San-Miguel-Ayanz, J., and Stibig, H.-J., Food and Agriculture Organization of the United Nations andEuropean Comission Joint Research Centre, Rome, FAO, 2012.
Farquhar, G. D. and Sharkey, T. D.: Stomatal Conductanceand Photosynthesis, Annu. Rev. Plant Phys., 33, 317–345,https://doi.org/10.1146/annurev.pp.33.060182.001533, 1982.
Farquhar, G. D., von Caemmerer, S., and Berry, J. A.:A biochemical model of photosynthetic CO2 assim-
422 F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro
ilation in leaves of C3 species, Planta, 149, 78–90,https://doi.org/10.1007/BF00386231, 1980.
Garten, J. C., Hanson, P., Todd, J. D., Lu, B., and Brice, D.: Natural15N and 13C abundance as indicators of forest N status and soil Cdynamics, in: Stable Isotopes in Ecology and Environmental Sci-ence, second ed., edited by: Michener, R., Lajtha, K., BlackwellScience, Oxford, 61–82, 2008.
Garten, C. T., Classen, A. T., and Norby, R. J.: Soil mois-ture surpasses elevated CO2 and temperature as a control onsoil carbon dynamics in a multi-factor climate change exper-iment, Plant Soil, 319, 85–94, https://doi.org/10.1007/s11104-008-9851-6, 2009.
Gerschlauer, F. and Kiese, R.: SP3_d15N/d13C natural abun-dance, TN, and SOC, https://www.kilimanjaro.biozentrum.uni-wuerzburg.de/Data/Data.aspx, last access: 31 August 2016.
Gerschlauer, F., Dannenmann, M., Kühnel, A., Meier, R., Kolar, A.,Butterbach-Bahl, K., and Kiese, R.: Gross Nitrogen Turnover ofNatural and Managed Tropical Ecosystems at Mt. Kilimanjaro,Tanzania, Ecosystems, 19, 1–18, https://doi.org/10.1007/s10021-016-0001-3, 2016.
Gruber, N. and Galloway, J. N.: An Earth-system perspec-tive of the global nitrogen cycle, Nature, 451, 293–296,https://doi.org/10.1038/nature06592, 2008.
Gütlein, A., Zistl-Schlingmann, M., Becker, J. N., Cornejo,N. S., Detsch, F., Dannenmann, M., Appelhans, T., Her-tel, D., Kuzyakov, Y., and Kiese, R.: Nitrogen turnoverand greenhouse gas emissions in a tropical alpine ecosys-tem, Mt. Kilimanjaro, Tanzania, Plant Soil, 411, 243–259,https://doi.org/10.1007/s11104-016-3029-4, 2017.
Gütlein, A., Gerschlauer, F., Kikoti, I., and Kiese, R.: Impacts of cli-mate and land use on N2O and CH4 fluxes from tropical ecosys-tems in the Mt. Kilimanjaro region, Tanzania, Glob. ChangeBiol., 24, 1239–1255, https://doi.org/10.1111/gcb.13944, 2018.
Hedin, L. O., Brookshire, E. N. J., Menge, D. N. L., andBarron, A. R.: The Nitrogen Paradox in Tropical For-est Ecosystems, Annu. Rev. Ecol. Evol. S., 40, 613–635, https://doi.org/10.1146/annurev.ecolsys.37.091305.110246,2009.
Hemp, A.: Climate change-driven forest fires marginalize the im-pact of ice cap wasting on Kilimanjaro, Glob. Change Biol., 11,1013–1023, https://doi.org/10.1111/j.1365-2486.2005.00968.x,2005.
Hickman, J. E., Palm, C. A., Mutuo, P., Melillo, J. M., andTang, J.: Nitrous oxide (N2O) emissions in response to in-creasing fertilizer addition in maize (Zea mays L.) agricul-ture in western Kenya, Nutr. Cycl. Agroecosys., 100, 177–187,https://doi.org/10.1007/s10705-014-9636-7, 2014.
Högberg, P.: 15N natural abundance in soil-plant systems,Tansley Review No. 95, New Phytol., 137, 179–203,https://doi.org/10.1046/j.1469-8137.1997.00808.x, 1997.
Houlton, B. Z. and Bai, E.: Imprint of denitrifying bacteria on theglobal terrestrial biosphere, P. Natl. Acad. Sci. USA, 106, 21713–21716, https://doi.org/10.1073/pnas.0912111106, 2009.
Kirschbaum, M. U. F.: The temperature dependence of soil or-ganic matter decomposition, and the effect of global warmingon soil organic C storage, Soil Biol. Biochem., 27, 753–760,https://doi.org/10.1016/0038-0717(94)00242-S, 1995.
Kohn, M. J.: Carbon isotope compositions of terrestrialC3 plants as indicators of (paleo)ecology and (pa-
leo)climate, P. Natl. Acad. Sci. USA, 107, 19691–19695,https://doi.org/10.1073/pnas.1004933107, 2010.
Körner, C., Farquhar, G. D., and Wong, S. C.: Carbon Isotope Dis-crimination by Plants Follows Latitudinal and Altitudinal Trends,Oecologia, 88, 30–40, 1991.
Leirós, M. C., Trasar-Cepeda, C., Seoane, S., and Gil-Sotres,F.: Dependence of mineralization of soil organic matter ontemperature and moisture, Soil Biol. Biochem., 31, 327–335,https://doi.org/10.1016/S0038-0717(98)00129-1, 1999.
Lewis, S. L., Lloyd, J., Sitch, S., Mitchard, E. T. A., andLaurance, W. F.: Changing Ecology of Tropical Forests: Ev-idence and Drivers, Annu. Rev. Ecol. Evol. S., 40, 529–549, https://doi.org/10.1146/annurev.ecolsys.39.110707.173345,2009.
Ma, J.-Y., Sun, W., Liu, X.-N., and Chen, F.-H.: Variation in the Sta-ble Carbon and Nitrogen Isotope Composition of Plants and Soilalong a Precipitation Gradient in Northern China, PLOS ONE, 7,e51894, https://doi.org/10.1371/journal.pone.0051894, 2012.
Majule, A. E.: Impacts of Land Use/Land Cover Changes on SoilDegradation and Biodiversity on the Slopes of Mount Kiliman-jaro, Tanzania, LUCID Project, International Livestock ResearchInstitute, 2003.
Mariotti, A., Germon, J. C., Hubert, P., Kaiser, P., Letolle, R.,Tardieux, A., and Tardieux, P.: Experimental determination of ni-trogen kinetic isotope fractionation: some principles; illustrationfor the denitrification and nitrification processes, Plant Soil, 62,413–430, 1981.
Martinelli, L. A., Piccolo, M. C., Townsend, A. R., Vitousek, P. M.,Cuevas, E., McDowell, W., Robertson, G. P., Santos, O. C., andTreseder, K.: Nitrogen stable isotopic composition of leaves andsoil: Tropical versus temperate forests, Biogeochemistry, 46, 45–65, https://doi.org/10.1007/BF01007573, 1999.
Mganga, K. Z. and Kuzyakov, Y.: Glucose decomposition and its in-corporation into soil microbial biomass depending on land use inMt. Kilimanjaro ecosystems, Eur. J. Soil Biol., 62, 74–82, 2014.
Michener, R. H. and Lajtha, K. (Eds.): Stable isotopes in ecologyand environmental science, 2nd ed., Ecological methods and con-cepts series, Blackwell Pub, Malden, MA, 2007.
Nardoto, G. B., Quesada, C. A., Patiño, S., Saiz, G., Baker, T.R., Schwarz, M., Schrodt, F., Feldpausch, T. R., Domingues,T. F., Marimon, B. S., Junior, B.-H. M., Vieira, I. C. G.,Silveira, M., Bird, M. I., Phillips, O. L., Lloyd, J., andMartinelli, L. A.: Basin-wide variations in Amazon forestnitrogen-cycling characteristics as inferred from plant and soil15N : 14N measurements, Plant Ecol. Divers., 7, 173–187,https://doi.org/10.1080/17550874.2013.807524, 2014.
Ometto, J. P. H. B., Ehleringer, J. R., Domingues, T. F., Berry,J. A., Ishida, F. Y., Mazzi, E., Higuchi, N., Flanagan, L. B.,Nardoto, G. B., and Martinelli, L. A.: The stable carbon andnitrogen isotopic composition of vegetation in tropical forestsof the Amazon Basin, Brazil, Biogeochemistry, 79, 251–274,https://doi.org/10.1007/s10533-006-9008-8, 2006.
Ortiz, C., Vázquez, E., Rubio, A., Benito, M., Schindlbacher, A.,Jandl, R., Butterbach-Bahl, K., and Díaz-Pinés, E.: Soil organicmatter dynamics after afforestation of mountain grasslands inboth a Mediterranean and a temperate climate, Biogeochem-istry, 131, 267–280, https://doi.org/10.1007/s10533-016-0278-5,2016.
F. Gerschlauer et al.: Stable isotopes of C and N at Mount Kilimanjaro 423
Pan, Y., Birdsey, R. A., Fang, J., Houghton, R., Kauppi, P. E., Kurz,W. A., Phillips, O. L., Shvidenko, A., Lewis, S. L., Canadell,J. G., Ciais, P., Jackson, R. B., Pacala, S. W., McGuire, A. D.,Piao, S., Rautiainen, A., Sitch, S., and Hayes, D.: A Large andPersistent Carbon Sink in the World’s Forests, Science, 333, 988–993, https://doi.org/10.1126/science.1201609, 2011.
Panettieri, M., Rumpel, C., Dignac, M. F., and Chabbi, A.: Doesgrassland introduction into cropping cycles affect carbon dynam-ics through changes of allocation of soil organic matter withinaggregate fractions?, Sci. Total Environ., 576, 251–263, 2017.
Pardo, L. H. and Nadelhoffer, K. J.: Using Nitrogen Isotope Ratiosto Assess Terrestrial Ecosystems at Regional and Global Scales,in: Isoscapes, edited by: West, J. B., Bowen, G. J., Dawson, T.E., and Tu, K. P., Springer, the Netherlands, Dordrecht, 221–249,2010.
Peterson, B. J. and Fry, B.: Stable Isotopes in Ecosys-tem Studies, Annu. Rev. Ecol. Syst., 18, 293–320,https://doi.org/10.1146/annurev.es.18.110187.001453, 1987.
Piccolo, M. C., Neill, C., Melillo, J. M., Cerri, C. C., andSteudler, P. A.: 15N natural abundance in forest and pasturesoils of the Brazilian Amazon Basin, Plant Soil, 182, 249–258,https://doi.org/10.1007/BF00029056, 1996.
R Core Team: R: A language and environment for statistical com-puting, R Foundation for Statistical Computing, Vienna, Austria,2015.
Robinson, D.: δ15N as an integrator of the nitrogen cycle,Trends Ecol. Evol., 16, 153–162, https://doi.org/10.1016/S0169-5347(00)02098-X, 2001.
Saiz, G. and Albrecht, A.: Methods for smallholder quantification ofsoil carbon stocks and stock changes, in: Measurement methodsStandard Assessment Of Agricultural Mitigation Potential AndLivelihoods (SAMPLES), edited by: Rosenstock, T. S., Rufino,M. C., Butterbach-Bahl, K., Wollenberg, E., and Richards, M.,ISBN 978-3-319-29792-7, CGIAR Research Program on Cli-mate Change, Agriculture and Food Security, 135–162, 2016.
Saiz, G., Bird, M., Domingues, T., Schrodt, F., Schwarz, M.,Feldpausch, T., Veenendaal, E., Djagbletey, G., Hien, F., Com-paore, H., Diallo, A., and Lloyd, J.: Variation in soil car-bon stocks and their determinants across a precipitation gra-dient in West Africa, Glob. Change Biol., 18, 1670–1683,https://doi.org/10.1111/j.1365-2486.2012.02657.x, 2012.
Saiz, G., Bird, M., Wurster, C., Quesada, C. A., Ascough, P.,Domingues, T., Schrodt, F., Schwarz, M., Feldpausch, T. R.,Veenendaal, E., Djagbletey, G., Jacobsen, G., Hien, F., Com-paore, H., Diallo, A., and Lloyd, J.: The influence of C3 and C4vegetation on soil organic matter dynamics in contrasting semi-natural tropical ecosystems, Biogeosciences, 12, 5041–5059,https://doi.org/10.5194/bg-12-5041-2015, 2015a.
Saiz, G., Wynn, J. G., Wurster, C. M., Goodrick, I., Nelson, P. N.,and Bird, M. I.: Pyrogenic carbon from tropical savanna burning:production and stable isotope composition, Biogeosciences, 12,1849–1863, https://doi.org/10.5194/bg-12-1849-2015, 2015b.
Saiz, G., Wandera, F. M., Pelster, D. E., Ngetich, W., Okalebo, J. R.,Rufino, M. C., and Butterbach-Bahl, K.: Long-term assessmentof soil and water conservation measures (Fanya-juu terraces) onsoil organic matter in South Eastern Kenya, Geoderma, 274, 1–9,https://doi.org/10.1016/j.geoderma.2016.03.022, 2016.
Saiz, G., Goodrick, I., Wurster, C., Nelson, P. N., Wynn, J., andBird, M.: Preferential Production and Transport of Grass-Derived
Schellenberger Costa, D., Gerschlauer, F., Pabst, H., Kühnel,A., Huwe, B., Kiese, R., Kuzyakov, Y., Kleyer, M., andKühn, I.: Community-weighted means and functional dis-persion of plant functional traits along environmental gra-dients on Mount Kilimanjaro, J. Veg. Sci., 28, 684–695,https://doi.org/10.1111/jvs.12542, 2017.
Smith, P., Clark, H., Dong, H., Elsiddig, E.A., Haberl, H., Harper,R., House, J., Jafari, M., Masera, O., Mbow, C., Ravindranath,N. H., Rice, C. W., Roble do Abad, C., Romanovskaya, A., Sper-ling, F., and Tubiello, F.: Chapter 11 – Agriculture, forestry andother land use (AFOLU), in: Climate Change 2014: Mitigation ofClimate Change, IPCC Working Group III Contribution to AR5,Cambridge University Press, 2014.
Soini, E.: Land use change patterns and livelihood dynamics on theslopes of Mt. Kilimanjaro, Tanzania, Agric. Syst., 85, 306–323,https://doi.org/10.1016/j.agsy.2005.06.013, 2005.
Sotta, E. D., Corre, M. D., and Veldkamp, E.: Differing N status andN retention processes of soils under old-growth lowland forest inEastern Amazonia, Caxiuanã, Brazil, Soil Biol. Biochem., 40,740–750, https://doi.org/10.1016/j.soilbio.2007.10.009, 2008.
Stevenson, B. A., Parfitt, R. L., Schipper, L. A., Baisden, W. T., andMudge, P.: Relationship between soil δ15N, C/N and N lossesacross land uses in New Zealand, Agr. Ecosyst. Environ., 139,736–741, https://doi.org/10.1016/j.agee.2010.10.020, 2010.
Swap, R. J., Aranibar, J. N., Dowty, P. R., Gilhooly, W. P., andMacko, S. A.: Natural abundance of 13C and 15N in C3 andC4 vegetation of southern Africa: patterns and implications,Glob. Change Biol., 10, 350–358, https://doi.org/10.1111/j.1365-2486.2003.00702.x, 2004.
Tieszen, L. L., Senyimba, M. M., Imbamba, S. K., and Troughton, J.H.: The distribution of C3 and C4 grasses and carbon isotope dis-crimination along an altitudinal and moisture gradient in Kenya,Oecologia, 37, 337–350, https://doi.org/10.1007/BF00347910,1979.
Townsend, A. R., Cleveland, C. C., Houlton, B. Z., Alden,C. B., and White, J. W.: Multi-element regulation of thetropical forest carbon cycle, Front. Ecol. Environ., 9, 9–17,https://doi.org/10.1890/100047, 2011.
Traoré, S., Ouattara, K., Ilstedt, U., Schmidt, M., Thiombiano,A., Malmer, A., and Nyberg, G.: Effect of land degradationon carbon and nitrogen pools in two soil types of a semi-arid landscape in West Africa, Geoderma, 241–242, 330–338,https://doi.org/10.1016/j.geoderma.2014.11.027, 2015.
van der Merwe, N. J. and Medina, E.: Photosynthesis and 13C12Cratios in Amazonian rain forests, Geochim. Cosmochim. Ac.,53, 1091–1094, https://doi.org/10.1016/0016-7037(89)90213-5,1989.
van Reeuwijk, L. (Ed.): Procedures for Soil Analysis 9, TechnicalPaper, International Soil Reference and Information Centre, Wa-geningen, 2002.
Vitousek, P. M.: Litterfall, Nutrient Cycling, and NutrientLimitation in Tropical Forests, Ecology, 65, 285–298,https://doi.org/10.2307/1939481, 1984.
Wang, C., Houlton, B. Z., Liu, D., Hou, J., Cheng, W., and Bai, E.:Stable isotopic constraints on global soil organic carbon turnover,
Wang, L., D’Odorico, P., Ries, L., and Macko, S. A.: Pat-terns and implications of plant-soil δ13C and δ15N valuesin African savanna ecosystems, Quaternary Res., 73, 77–83,https://doi.org/10.1016/j.yqres.2008.11.004, 2010.
Wynn, J. G. and Bird, M. I.: C4-derived soil organic carbon decom-poses faster than its C3 counterpart in mixed C3/C4 soils, Glob.Change Biol., 13, 2206–2217, https://doi.org/10.1111/j.1365-2486.2007.01435.x, 2007.
Zaehle, S.: Terrestrial nitrogen–carbon cycle interactions atthe global scale, Philos. T. R. Soc. B, 368, 20130125,https://doi.org/10.1098/rstb.2013.0125, 2013.
Zech, M., Bimüller, C., Hemp, A., Samimi, C., Broesike, C.,Hörold, C., and Zech, W.: Human and climate impact on 15Nnatural abundance of plants and soils in high-mountain ecosys-tems: a short review and two examples from the Eastern Pamirsand Mt. Kilimanjaro, Isotopes Environ. Health Stud., 47, 286–296, https://doi.org/10.1080/10256016.2011.596277, 2011.
Zhou, Y., Fan, J., Zhang, W., Harris, W., Zhong, H., Hu,Z., and Song, L.: Factors influencing altitudinal patterns ofC3 plant foliar carbon isotope composition of grasslandson the Qinghai-Tibet Plateau, China, Alpine Bot., 121, 79,https://doi.org/10.1007/s00035-011-0093-5, 2011.
Zhu, Y., Jiang, Y., Liu, Q., Kang, M., Spehn, E. M., and Körner,C.: Elevational Trends of Biodiversity and Plant Traits Do NotConverge – A Test in the Helan Range, NW China, Plant Ecol.,205, 273–283, 2009.
